Back to Search
Start Over
Image registration for a UV–Visible dual-band imaging system
- Source :
- Optics and Lasers in Engineering. 105:209-218
- Publication Year :
- 2018
- Publisher :
- Elsevier BV, 2018.
-
Abstract
- The detection of corona discharge is an effective way for early fault diagnosis of power equipment. UV–Visible dual-band imaging can detect and locate corona discharge spot at all-weather condition. In this study, we introduce an image registration protocol for this dual-band imaging system. The protocol consists of UV image denoising and affine transformation model establishment. We report the algorithm details of UV image preprocessing, affine transformation model establishment and relevant experiments for verification of their feasibility. The denoising algorithm was based on a correlation operation between raw UV images, a continuous mask and the transformation model was established by using corner feature and a statistical method. Finally, an image fusion test was carried out to verify the accuracy of affine transformation model. It has proved the average position displacement error between corona discharge and equipment fault at different distances in a 2.5m-20 m range are 1.34 mm and 1.92 mm in the horizontal and vertical directions, respectively, which are precise enough for most industrial applications. The resultant protocol is not only expected to improve the efficiency and accuracy of such imaging system for locating corona discharge spot, but also supposed to provide a more generalized reference for the calibration of various dual-band imaging systems in practice.
- Subjects :
- 010302 applied physics
Image fusion
Computer science
business.industry
Mechanical Engineering
Image registration
Fault (power engineering)
01 natural sciences
Atomic and Molecular Physics, and Optics
Electronic, Optical and Magnetic Materials
010309 optics
Transformation (function)
Feature (computer vision)
0103 physical sciences
Calibration
Computer vision
Affine transformation
Artificial intelligence
Electrical and Electronic Engineering
business
Corona discharge
Subjects
Details
- ISSN :
- 01438166
- Volume :
- 105
- Database :
- OpenAIRE
- Journal :
- Optics and Lasers in Engineering
- Accession number :
- edsair.doi...........fa5a70b500b1a107f2c69ac528664fde